I've reviewed a paper recently.
In my review, I pointed out what I thought to be a major flaw in the approach proposed by the authors, and suggested a method that would circumvent that problem and probably not be much slower (it was important that this algorithm be real fast).
Normally, in that scenario, one of two things happen: either I misunderstood the approach, and the authors kindly point out that Reviewer 3 is out of her mind and shouldn't be allowed in a five-mile radius from science, or the paper ends up rejected, possibly after the authors devoted three-quarter of a page of their response to shabby arguments that try to justify why they won't try out any of the suggestions made by the reviewers.
I just got the response to the reviewers.
Something magical happened.
The authors took my comment seriously. They implemented the approach I was suggesting. They found out that it performed better than theirs (including in terms of runtime!). They admitted to it. They profusely thanked me for the suggestion. They removed their flaky algorithm from their paper and replaced it with mine.
It feels like seeing a unicorn pooping rainbows, or whichever Internet imagery you want to use.
Like, wow, science. As it should be done.
Cherry on top: it's an open access journal that also practices open peer review. While I was worried my comments would lead the authors to poison my beer at the next conference at which we meet, this actually mean that not only to they get a (slightly) better paper out of it, but I also still get some credit.
In the meanwhile, on FB, someone whose child has leukemia is blaming everyone from the government to researchers for the ordeal her little girl is going through.
I do what I can, okay.